ExpertsFMUber CEO Dara Khosrowshahi
Acquired · June 13, 2023
5 Selected Lessons:
heuristic · 2 framework · 2 priority · 1
Prioritize earner platform quality as the foundation for Uber's long-term win, above short-term financial optimization
Source
Acquired — Uber CEO Dara Khosrowshahi (June 13, 2023) · at 1:27:20 ▸ · at 1:30:40 ▸
When this applies
when a platform company must choose what to optimize for among competing priorities: financial metrics, consumer experience, or the welfare of the dependent supply-side population
The takeaway
Building the best platform for the supply-side population that depends on you is both a practical long-term competitive strategy and a values statement distinguishing a 'real world' tech company from one disconnected from ordinary people.
Why it works
A technology company that is uniquely connected to a broad, non-elite population (not just consumers or engineers) builds durable differentiation and long-term competitive position if it wins the trust and welfare of that population, because they are both the product's engine and underserved by typical tech-company practices.
Where it stops holding
This stops working when the supply-side population isn't large/dependent enough to be a meaningful differentiator, or when the practical business case for supply investment is weak.
Detail
hierarchy:
- Build the best platform/product experience for earners (drivers/couriers) — described as the foundation for winning long-term.
- Be a technology company connected to the broad population/'real world,' not just an elite audience.
- Turn around the business financially and build a strong team/service (acknowledged as already achieved).
rationale: If we build a company that has the best product and experience for earners, we're going to win long term... that's a company to be proud of.
In their words
There's a practical reality, which is if we build a company that has the best product and experience for earners, we're going to win long term. But if we're that technology company that's like very much connected, not with the elite, but with an earner base and the broad population
Supply-led growth for fragmented-supply marketplaces
Source
Acquired — Uber CEO Dara Khosrowshahi (June 13, 2023) · at 18:20 ▸ · at 22:48 ▸
When this applies
when building any two-sided marketplace where supply is fragmented and liquidity/conversion depends on density
The takeaway
In a fragmented-supply marketplace, prioritize investment in deepening local supply over broad demand/brand building, because supply density compounds into better conversion, pricing, and liquidity which pulls in demand.
Why it works
Building up supply density in a given market increases conversion and data for optimization (e.g., search ranking), which in turn attracts more demand without proportional demand-side spend — supply becomes self-reinforcing growth.
One level up
Go after the larger, more fragmented side of a market, because fragmentation is where a scaled aggregator can build durable advantage.
Where it stops holding
This stops working when supply is not meaningfully fragmented (e.g., a few dominant suppliers already control access), or when demand-side brand/trust is the actual bottleneck rather than density.
Detail
procedure:
- Identify which side of the marketplace is more fragmented (more individual suppliers) — that's the larger addressable opportunity.
- Invest in building up supply density market-by-market rather than primarily building demand/brand.
- Use increased supply density to improve conversion and ranking (e.g., via search/marketplace algorithms), which pulls in more organic demand.
- Exploit cross-side recruitment advantages where one vertical (e.g., food delivery) has a lower barrier to entry than another (e.g., ride-hailing), using it to onboard supply faster, then upsell.
benchmarks: {"example":"post-pandemic Uber focused on bringing drivers back to the platform and gained category share versus Lyft"}
rationale: If you have a hundred hotels in a market and you expand that to 200 hotels in a market, that market is also going to convert better — not only do you build a new segment of demand but convert existing demand better.
In their words
So I think one of the lessons is like, hey, go after the larger market. And if you're a marketplace business, go after fragmentation of supply, which is if you think about hotels, there's so many more hotels in the world than there are airlines.
Three-question test for evaluating a job opportunity
Source
Acquired — Uber CEO Dara Khosrowshahi (June 13, 2023) · at 46:12 ▸
When this applies
when evaluating any major career opportunity or role change
The takeaway
Evaluate opportunities on three independent axes: quality of people you'd work with, your personal ability to have impact, and the external significance of the organization's impact.
Why it works
These three questions separately test relational fit (team), personal agency (can you move the needle), and external significance (does the mission matter) — together they cover whether the role will be both fulfilling and impactful.
Where it stops holding
This stops working when one axis is so overwhelmingly negative (e.g., toxic team) that it should veto the decision regardless of the other two.
Detail
procedure:
- Ask: do you work with people whom you like and can learn from?
- Ask: can you, as an individual, make an impact?
- Ask: is the place or company you're at going to make an impact?
rationale: I wasn't sure number one, but I was a CEO so I could build my own team — the other two axes were clearly yes for Uber.
In their words
I always tell people, like, I look for three things, It's: do you work with people whom you like and you can learn from? Can you, as an individual, make an impact? And then, is the place or the company that you're at going to make an impact?
Option time-value, but demand a decision when asked
Source
Acquired — Uber CEO Dara Khosrowshahi (June 13, 2023) · at 4:44 ▸
When this applies
when you hold optionality in a negotiation but the uncertainty itself is damaging the underlying asset or team
The takeaway
Hold optionality as long as it costs you nothing, but recognize when ongoing uncertainty is itself the cost, and force a decision at that point.
Why it works
An option's value includes time value, so you should delay exercising it until the last useful moment; but prolonged ambiguity can itself destroy value for the party awaiting the decision, so forcing a yes/no at the right time converts paralysis into stability.
Where it stops holding
This stops working when forcing an early decision eliminates information that would have changed the outcome, or when the party being pressured has no real urgency to resolve.
Detail
rule: If you're the party whose business is destabilized by someone else's unexercised option, proactively ask them to decide rather than wait passively.
threshold: When instability itself (not just the eventual decision) is costing you, that's the signal to force resolution.
rationale: Companies thrive on certainty and rhythm; ambiguity about control or ownership is itself corrosive.
In their words
And at the time, you know, one of the values of an option is time value, right? You don't want to exercise an option before the last moment that you can.
Resist take-rate temptation — keep take rate low for durability over quick margin wins
Source
Acquired — Uber CEO Dara Khosrowshahi (June 13, 2023) · at 1:23:48 ▸
When this applies
when any intermediary/platform business faces the easy lever of raising its cut of transaction value to hit short-term financial targets
The takeaway
Resist the temptation to raise the platform's cut of transaction value to hit short-term financial targets; instead pursue durable margin expansion through operational efficiency, because a high take rate is 'a rake too far' that erodes the ecosystem you depend on.
Why it works
Raising take rate is 'pure margin dollars' with no corresponding value creation, which erodes the ecosystem partners' (drivers'/couriers') incentive to participate and ultimately threatens long-term volume growth; keeping take rate low requires hard, continuous operational efficiency work instead, but sustains durability.
One level up
Durability in a multi-sided ecosystem is built by leaving more value with ecosystem partners, even when short-term incentives push toward extracting more.
Where it stops holding
This stops working when the ecosystem has no viable alternative to your platform (true monopsony), in which case low take rate doesn't protect against defection, or when operational efficiency gains are genuinely exhausted and growth truly requires margin extraction.
Detail
rule: Deliberately resist raising take rate even when it's the easiest lever to hit quarterly numbers; pursue efficiency (fraud reduction, automation, cost discipline) instead to keep take rate low.
threshold: At Expedia, took take rate from 25% down to ~17% over 13 years; at Uber, last quarter gross bookings grew ~22% while driver/courier earnings (incl. tips) grew 30%, while still expanding margins and achieving positive free cash flow.
rationale: Any quarter I can deliver anything on the bottom line if I can move my take rate up a little bit. But like, it's too easy. It's too tempting... high take rates are dangerous.
In their words
watch every single nickel and dime, be incredibly efficient in everything that you do, automate everything, get fraud out of the system, et cetera, so that you can actually operate a business at scale at the lowest take rate possible.